DETEKSI ALFABET BISINDO MENGGUNAKAN MEDIAPIPE HOLISTIC SECARA REAL-TIME

Abstract

The resulting system will detect and classify gestures or movements that speak the B language contained in it directly using a holistic mediapipe to detect the skeleton on the hands, body, and face and use a landmark model of poses, faces and each to generate a total of 543 landmarks. (33 pose landmarks, 468 facial landmarks, and 21 hand landmarks).